Atomic Earthworms

A group of materials scientists has unleashed atomic-scale earthworms, which eat meandering trenches just a few atoms wide through a semiconductor. Spawned when the researchers mixed two elements in a thin film and let it cool, the worms might one day serve to mass-produce chip patterns so tiny that they would alter the quantum behavior of the electrons they confine. Such "quantum wires" could lead to improved sensing devices or lasers.

Semiconductor lasers like those in CD players already exploit quantum mechanics by trapping electrons in an atoms-thick layer called a quantum well. By the laws of quantum physics, the close confinement only allows the electrons a few distinct energy states. Because a laser works by forcing electrons to jump between energy states, better confinement translates to a more efficient laser--one that fits in your living room instead of a physics lab. Now materials scientists are dreaming of the next step in electron confinement--quantum wires, which would trap electrons in one dimension and lead to even more efficient lasers, for example.

"The trick," says Mohan Krishnamurthy, a materials engineer at Michigan Technological University in Houghton, is "to come up with something that produces high-quality wires in an elegant and relatively high-throughput manner." So Krishnamurthy tried to create conditions in which the wires would assemble themselves. As he and his colleagues report in this week's Physical Review Letters, they started by producing a thin film from an alloy of germanium and tin--two semiconductors that do not like to mix. Predictably, the tin separated out as the mixture cooled, but what happened next as the tin sought to lower its energy was unexpected. "Like an earthworm, the globs of tin eat up the [alloy], spit out the germanium, and keep the tin," Krishnamurthy says. In their wake they left wiggly trenches of germanium.

An obvious problem remains before the trenches--or their walls--can be turned into mass-produced quantum wires: controlling the meandering tin droplets so they form straighter paths. "The chance of this particular thing working is zero," says Max Lagally, a materials scientist at the University of Wisconsin, Madison. But if you could control the density of the tin droplets and the direction they travel, he says, the approach "could be useful" for devices that don't require a precise wiring pattern--such as infrared detectors.